A weighted prediction-based selection criterion for response surface designs
Version of Record online: 25 JUL 2011
Copyright © 2011 John Wiley & Sons, Ltd.
Quality and Reliability Engineering International
Special Issue: ENBIS 10
Volume 27, Issue 5, pages 719–729, July 2011
How to Cite
Emmett, M., Goos, P. and Stillman, E. C. (2011), A weighted prediction-based selection criterion for response surface designs. Qual. Reliab. Engng. Int., 27: 719–729. doi: 10.1002/qre.1231
- Issue online: 25 JUL 2011
- Version of Record online: 25 JUL 2011
- design of experiments;
- compound design criteria;
- logistic regression
The goal of response surface designs is typically to make precise predictions. A commonly used prediction-based design selection criterion isitV-optimality, which seeks designs that minimize the average prediction variance over the entire experimental region. We propose an alternative criterion, which seeks designs that yield small prediction variances particularly in those parts of the experimental region where a response is expected to be interesting, important, or desirable. The new criterion is a weighted V-optimality criterion, which attaches higher weights to areas with such interesting outcomes. The weights in the new criterion are derived from a logistic regression model. We illustrate the value of the new criterion using an example from the automotive industry. Copyright © 2011 John Wiley & Sons, Ltd.